The subsidy control provisions of the UK-EU trade and cooperation agreement: a framework for a new UK domestic subsidy regime – by George Peretz QC

The provisions of the (draft) UK-EU Trade and Cooperation Agreement (the “Agreement”) that govern subsidy control (Title XI, Chapter three, starting at page 184) are unprecedented in comparison with other free trade agreements.  They impose significant constraints on UK policy towards public subsidies.

What “subsidies” are caught by the Agreement?

As to scope, the definition of “subsidy” in Article 3.1 of the Agreement essentially parallels the definition of “State aid” in EU law.  Thus, it covers: public grants, loans, and guarantees; forgoing revenue otherwise due (e.g. tax waivers or debt write-offs); and the provision of goods or services or the purchase of goods and services, such as agreeing to purchase services at an overvalue or only in order to assist an operator. The provision of general infrastructure is not in terms excluded – as it is under Article 1.1(a)(1)(iii) of the WTO Agreement of Subsidies and Countervailing Measures – perhaps because it was considered that the provision of general infrastructure would not be a specific advantage, just as it is not a selective advantage under EU State aid law. 

A subsidy must also confer an advantage and be “specific” (a term derived from WTO subsidy law but which, both there and here, means much the same as “selective” in EU law). The assistance must also have actual or potential effects on trade and investment between the parties: whether that threshold is much higher than the notoriously low threshold that similar words have created in EU State aid law remains to be seen (it may be noted that Article 3 refers elsewhere to “material” effects on trade, perhaps suggesting that the omission of “material” at this point is deliberate). 

Tax measures can be subsidies.  Article 2(a) attempts to limit the application of the concept of “subsidy” to tax measures by setting out circumstances in which a general tax measure will not be considered “specific”.  These broadly mirror existing case-law of the ECJ in EU law: a tax measure will not be a subsidy unless some economic actors pay less tax than they would in comparison with a “normal” tax regime (the often-criticised “reference framework” concept in EU law); those economic actors must be treated more advantageously than those in a “comparable position” from the point of view of the objective and features of the tax regime; and favourable treatment is not a subsidy if it is justified by features such as the need to fight tax evasion, tax neutrality, the progressive or redistributive nature of a tax or its link to ability to pay (all of which mirrors current ECJ case-law on the nature and logic of the tax system).  So the subsidy provisions are more or less as likely to apply to tax measures as are EU rules on State aid (either to tax rulings, such as the Apple case (Cases T‑778/16 and T‑892/16), or to general tax measures that operate in favour of particular operators, such as the recent UK case on the tax treatment of certain controlled foreign companies).

Article 3.2 of the Agreement limits the scope of what counts as a “subsidy”.  Again, much of this mirrors EU State aid rules.  Subsidies to compensate damage caused by natural disasters or other exceptional non-economic occurrences (such as pandemics) are largely excluded from the rules on subsidies – but State aid falling in that category has to be cleared by the Commission under Article 107(2) of the Treaty on the Functioning of the EU, so, again, there is little change from the State aid position.  The exclusion of subsidies targeted at consumers also mirrors EU State aid law, as does the de minimis provision (set, at SDR 325,000, or ~EUR 380,000, over three years, at about twice as high as the EU de minimis threshold of EUR 200,000 over three years). 

There are, however, two significant departures from EU State aid law.  Subsidies to the audio-visual sector are entirely excluded,  so the rules do not apply to the funding of the BBC or to support for the film industry.  And the provisions of the agreement that prohibit subsidies or require remedial measures do not apply to subsidies granted on a temporary basis to respond to a national or global economic emergency, though such subsidies must be targeted, proportionate and effective.

Article 3.3 of the Agreement again rings bells for State aid lawyers by excluding from the rules subsidies granted to providers of services of “public economic interest” (cf “general economic interest” in EU law) if application of the rules would obstruct them in performing their tasks: Article 3.3.2 requires the avoidance of over-compensation and cross-subsidy, while Article 3.3.4 fixes a higher de minimis threshold for such subsidies.  Again, all familiar to State aid practitioners.

What rules apply to subsidies caught by the Agreement?

Articles 3.4.2 and 3.5 of the Agreement provide that certain types of subsidies must be prevented by the parties, if they are such as to have an actual or potential “material” effect on trade or investment between the parties.  They are:

  • unlimited state guarantees;
  • rescue and restructuring assistance where there is no credible restructuring plan, no provision for current owners to contribute to the plan (except for SMEs), or no social hardship or severe market failure consequent on business failure (this provision, and the first bullet above, parallel provisions in the UK/Japan FTA);
  • assistance to banks, credit institutions and insurance companies exceeding what is needed to secure an orderly exit from the market unless there is a credible restructuring plan;
  • export subsidies other than short-term credit insurance for non-marketable risks;
  • subsidies conditional on domestic content; and
  • subsidies to air carriers for the operation of routes, unless they are start-ups, where there is a public service obligation, or in special cases where the funding provides societal benefits.

Article 3.5 also creates special rules for assistance for secure and sustainable energy and for environmental sustainability, and for large cross border or international cooperation projects.

Outside those categories, the Agreement commits the parties to applying the general principles set out in Article 3.4.1 to all subsidies within the scope of the Agreement.  Those principles are familiar to any student of Commission State aid policy: assistance should pursue a specific public policy objective to meet an identified market failure or address an “equity rationale”; should be proportionate and necessary; should generate a change in economic behaviour that assists in achieving the objective; should not compensate for costs the beneficiary would have funded itself; that there is no less distortive alternative; and that positive effects outweigh negative ones.

In applying the Article 3.4.1 principles, the Joint Declaration on Subsidy Control Policies provides at page 5 useful guidance on some issues. First, it sets out a framework for clearance of subsidies to address regional disadvantage – important to any ‘levelling up’ policy. The framework itself is fairly anodyne (it requires policy to take account of the needs of the area, the size of the beneficiary and the size of the project, and to require a substantial contribution from the beneficiary), but it means that there is no doubt that the UK can develop its own system of what State aid lawyers think of as regional aid. It also makes it clear that such assistance cannot have as its main purpose or effect the poaching of activity from the other side – thereby discouraging subsidy races between the EU and UK (something that is very much in the UK interest, given the greater resources of many EU Member States). Further provisions of the Declaration deal with transport (generally summarising standard EU State aid policy and law on subsidies to airports, road infrastructure and ports) and research and development (again summarising the established EU law and policy approach). The value of these latter declarations is largely to make it clear that the UK government, committed to major investment in both transport and R&D, can do so within the framework of the Agreement – though it will be observed that it could also have done so within the framework of EU State aid rules.

Article 3.6 commits the parties to ensuring that beneficiaries only use assistance for the purpose for which it was granted (no “misuse of aid”, in EU law).

How are these rules to be enforced internally, within the UK?

The provisions do not commit the UK to replicate the ex ante mechanisms of EU State aid law (i.e. that State aid may not be implemented unless and until it is notified to and cleared by the European Commission). Note, though, that most State aid actually granted in the EU is granted under legislative block exemptions that mean it does not have to be notified.

One important commitment – especially bearing in mind concerns expressed recently over Covid-19 contracts – is a commitment to transparency in Article 3.7 of the Agreement.  All subsidies have to be published within six months with details and, if there is a letter before action from an interested party, that party has to be provided with all relevant information within 28 days (see Article 3.7.5).

But, more substantively, the UK is committed, under Articles 3.9, 3.10, and 3.11 respectively, to:

  • an independent authority or body that will play “an appropriate role” in its regime;
  • a power for courts to review compliance with the principles set out above by granting authorities and to review decisions of the independent authority, to do so on application by interested parties with standing, and to grant remedies including injunctions and orders to recover assistance granted; and
  • a power of recovery if an application is made in time (there are complex provisions as to what “in time” means) by an interested party on the ground that a measure should have been treated as a subsidy but wasn’t, or has misapplied the principles applicable to that subsidy.

In deference to the principle of Parliamentary sovereignty, recovery is not required where the subsidy is granted on the basis of an Act of the UK Parliament (Article 3.11.5),  but, importantly, that exception does not cover subsidies granted by the devolved Parliaments.

Against that background, the UK Government’s resistance to an ex ante commitment may well prove not to matter very much.  No beneficiary (or provider of loan or equity finance) is going to commit to a project if it is vulnerable to being overturned by a court.  In some cases, certainty and immunity from challenge could be achieved by simply allowing the time for challenge to lapse before proceeding with a project; however, in others that will be too slow and uncertain a device.  In practice, therefore, the UK regime is going to have to provide some mechanism for clearance by the independent authority – a clearance that will have binding effect.  Once in place, any such mechanism is likely to have to be gone through before beneficiaries and funders will agree to proceed – effectively requiring an ex ante clearance as a matter of commercial necessity rather than law.  Moreover, it is hard to see that the Government would in reality be content with a system that left the courts as the judges of whether the Article 3.4.1 principles had been correctly applied, given the need for consistent and transparent policy so as to guide granting authorities and beneficiaries in working out what types of assistance are likely to be acceptable given the broad parameters of the Article 3.4.1 principles.  In practice, therefore, the application of the Article 3.4.1 principles is likely to have to be a matter for the independent authority rather than the courts. 

All this therefore points to a regime that in substance, if not in form, ends up rather similar to the structure under the EU State aid rules: an independent authority issuing policy guidance and deciding the application of the Article 3.4.1 principles in particular cases, and whose decision is in commercial reality if not in law needed before most significant projects involving subsidies can proceed, combined with a role for the courts in dealing with cases where there is a real argument about whether the measure is a subsidy at all and where the independent body has never been approached, and in reviewing the decisions of the independent body. 

How are these rules to be enforced by the EU, if it considers the UK rules have not worked?

All the following provisions apply equally, vice versa, where the UK is unhappy with an EU subsidy.

First, the EU can appear as an intervening party in any court action in the UK concerning the subsidy rules: Article 3.10.2.

More importantly, in relation to an individual case, the EU can seek information and consultations: Article 3.12 of the Agreement.  If that does not work, it can take remedial measures if there is evidence that a UK subsidy will cause, or runs a serious risk of causing, a significant negative effect on trade an investment as between the UK and the EU.  Article 3.12.5 and .6 emphasise that that assessment must be based on facts and not speculation, and on reliable evidence.  The remedial action must be limited to what is strictly necessary and proportionate to remedy the significant effect.  If the UK considers that the EU has exceeded its rights to take remedial measures, then it can take the matter to an arbitration panel. Failure to comply with that panel’s ruling (if it is against the EU) triggers a UK right to take remedial action.

A more general UK failure to comply with the subsidy control rules (e.g. a claim that UK law does not meet the requirements imposed by Article 3) gives rise (except in relation to Articles 3.9 and 3.10) to a dispute settlement procedure under the general dispute settlement rules in Part Six, Title I.

Article 10 of the NI Protocol

Article 3 of the Agreement makes no reference whatsoever to Article 10 of the Ireland/Northern Ireland Protocol (the “Protocol”).  But Article 10 – which applies the EU State aid rules in full to the UK in relation to measures that have an actual or potential effect on trade in goods between Northern Ireland and the EU – will apply to many UK measures which have effects in Northern Ireland.  Article 10 of the Protocol will apply in parallel with the UK regime set up to implement the subsidy provisions of the Agreement.  That is well short of ideal, as there will be two separate subsidy regimes applicable to the same measures (as to which, as Lady Bracknell might have said, to have one set of subsidy rules to worry about may be necessary: to have two looks like carelessness).  The only way of avoiding that would be to carve out measures subject to Article 10 of the Protocol from the UK regime; however, given the penumbra of uncertainty as to when Article 10 will apply, that is not a satisfactory option either.

As I explained in my blog post here, the problems with the application of Article 10 of the Protocol – and in particular the “reach-back” issue, that is to say, the likelihood of its applying to general UK measures affecting Great Britain as well as Northern Ireland, as well as to GB measures with an indirect effect in Northern Ireland – were not adequately addressed by the 8 December statement by the Joint Committee.  It is also unlikely to be assisted by guidance issued by the Secretary of State under section 48 of the Internal Market Act 2020: as I pointed out in my blog post on the Joint Committee statement, the ambit of Article 10 is for the courts, and ultimately the ECJ to determine: what the Joint Committee says, and what the Secretary of State says, is in the end no more than a view.

It is regrettable that the UK Government did not seek (or if it sought did not obtain) an amendment to Article 10 to reflect its commitments under the subsidy control provisions of the Agreement.  The result is going to be, inevitably, something of a mess.

The short-term and the medium-term

Article 3 of the Protocol requires implementing legislation by the UK Government.  The Sewel Convention is not in play as subsidy control is now a reserved matter outside devolved competence: section 52 of the Internal Market Act 2020.

It is, however, inconceivable that a new regime could be up and running by 1 January 2021 – less than a week away.  As far as I am aware, there is no detailed legislation prepared, and the sudden production of detailed new legislation would both be unwise at this hyper-late stage and contrary to the spirit if not the letter of section 53 of the Internal Market Act 2020 (which requires the devolved governments to be provided with a pre-publication draft of the Government’s response to its consultation on a new UK subsidy regime).

There therefore seems to me to be no real alternative to preserving the EU State aid rules, modified only to make sense outside the EU, until a new system is up and running.  Fortunately, in best “Blue Peter” fashion, draft legislation – in the form of a “no deal” statutory instrument under the EU (Withdrawal) Act 2018 prepared under the May Government – already exists.  With a few tweaks to deflect accusations that the current Government was doing anything designed by its Conservative predecessor, it could be pressed into service, and the recently-made statutory instrument removing the State aid rules from retained EU law could simply and quietly be revoked. It would also be sensible – since there will not be time to devise new ones – to press back into service all the current EU block exemption regulations, compliance with which will provide certainty that a subsidy is lawful.

What the current Government’s plans are in that regard will presumably be made clear when the Bill implementing the Agreement is published and voted on this week.


This blog post is no more than an immediate reaction to the detailed provisions in the Agreement. 

What can be said, however, is that the UK has accepted a framework that fundamentally resembles the EU regime, even though in form it self-consciously avoids expressly drawing on it.  The UK will have to create a robust domestic framework for the regulation of subsidies, including an independent regulator and a set of rights to bring subsidy control issues to the courts.  The challenge will be to create a regime which makes use of the ability to move away from some of the more rigid aspects of the EU State aid regime while continuing to provide a useful check and review on the power to subsidise – a power which can often be used wisely to promote very important public policy objectives, but can also be abused for bad political or even venal reasons, and in ways that seriously damage the legitimate interests of trading partners.


Monckton Chambers, London, and Law Library, Dublin

(This post originally appeared on

This entry was posted in EU/UK Trade and Cooperation Agreement, New UK subsidy control regime. Bookmark the permalink.